5. The experience of water development

5.1 The popularity of water development
5.2 Water development in the past
5.3 Negative aspects of water development - Technical problems
5.4 Administrative and financial problems
5.5 Environmental problems and land use
5.6 The question of equity

5.1 The popularity of water development

In arid and semi-arid areas the development of new water supplies has been, and for the most part still is, overwhelmingly the form of development most sought after by pastoralists. In a few cases this enthusiasm has been tempered by experience. In southeast Ethiopia in the early 1970s pastoralists said:

'Put in wells a long way from here otherwise people will come and the grass will be finished... All we need is enough (water) for our own people and we don't mind walking for a day to get it.' (Cossins, 1971a, p.45).

In Niger in 1971/72, in roughly the same area as the example of Tuareg seminomads used in Chapter 3 - an area where a number of factors including government water development had caused the collapse of the previous land-use system nearly three quarters (73%) of livestock owners surveyed wanted a further increase in the number of open wells, although somewhat less than a half of them (43%) wanted a further increase in the number of boreholes (Marty, 1972, p. 91). In some cases pastoralists in this region requested the government administration to close boreholes already in operation because of the disruption they cause (Bernus, 1981, p. 429). In many, even most, pastoral areas, however, enthusiasm for new water development remains strong. We shall return subsequently to the question of which pastoralists, within a group or area, are most enthusiastic and receive the most benefit from water development, and which the least.

5.2 Water development in the past

Water development in the past has undoubtedly had a very important impact on livestock output from the dry (arid and semi-arid) regions of Africa. Although there must have been some effect in terms of a reduction in the amount of energy wasted in trekking to water, it seems probable that the main impact has been through opening up for more intensive use areas which, prior to water development, could only be used for short periods in the year and by a few livestock, or which could not be used at all. For example, between 1965 and 1976 the area of land in Botswana accessible to domestic livestock approximately doubled as a consequence of borehole drilling programmes financed from both public and private sources (Sandford, 1977, p.B. 16). In Sudan the number of watering sites for livestock was allegedly quadrupled between 1957 and 1968 by constructing nearly 1000 water yards equipped with boreholes, reservoirs and dams (Ministry of Agriculture and National Council for Research, 1976). In both countries these measures were accompanied by tremendous increases in the livestock population able to take advantage of the feed resources made accessible by these developments. According to government estimates the cattle population in Botswana increased by 73% between 1965 and 1975 (Sandford, 1977, p.25) and in Sudan by 55% between 1957 and 1968 (FAO Production Yearbooks). As is usually the case with such statistics, not much confidence can be put in these figures.

5.3 Negative aspects of water development - Technical problems

In spite of the popularity of water development among pastoralists in dry regions and of the impressive scale on which it has taken place in some countries, there have also been several negative aspects to it. In many cases the technical efficiency of construction, operation and maintenance has been low, and a few years after construction many water points are no longer in use. In Kenya in 1976 only 12 out of 54 boreholes drilled in the northeast after 1969 were still operating (Kenya Ministry of Water Development, internal reports). In Botswana 40% of boreholes drilled never operate, and in a survey of the Botswana Western Sandveld region only 65% of water points (85% of these were boreholes) were found to be operating at the time of the survey, 19% had been completely abandoned²², and 16% were not functioning for more temporary reasons - usually mechanical failure of some kind (Hitchcock, 1978, pp. 143-157). In Niger out of 23 boreholes installed between 1961 and 1969 about 15 (65%) were still functioning by 1969 (Bernus, 1977, pp. 56-57). In Sudan, out of 145 boreholes in one area of Southern Darfur Province, 28 had broken or malfunctioning pumps in the mid-1970s and 44, although constructed, had never operated (Huntings Technical Services and Sir M. Macdonald and Partners, 1976, pp. 13 and 61). In one part of southeast Ethiopia in 1974 only 9 out of at least 17 boreholes known to have been drilled were still functioning (Livestock and Meat Board, 1974, Annex II).

22 Although this is not expressly stated it seems that boreholes which had never operated, because 'cry' or otherwise unsuccessful at first drilling (i.e. the 40% already referred to), are excluded from these Western Sandveld figures.

Boreholes probably have a worse performance than other types of water source but other kinds also have problems. Of 100 hafirs or hafir dams built in the northeast of Kenya in the 1970s 'many are silted in, some completely' (Axin et al, 1979, pp. 29 and 55) due to heavy rainfall causing unexpectedly high erosion, unrestricted access to the actual dam and side wells by livestock and damage consequent on this, and the absence of maintenance. In 1977 someone who had known Tanzania Maasailand for over 20 years commented:

'Most of the former permanent water supplies of Maasailand such as boreholes, dams and improved spring catchments and water pipes are either broken down, clogged up, working at reduced capacity or in need of hardware and repair to permit them to work effectively.' (Jacobs, 1977).

A large range development project in this area in the early 1970s constructed or rebuilt about 25 major dams (USAID/Tanzania Government, 1977, pp. 18-20).

'Unusually heavy rains in the spring of 1977 breached and destroyed all of the earthen dams constructed during the previous 7 years, all of which - according to the new hydrologist - had been built without proper subsoil and watershed assessment and without adequate wetting or compaction.' (Jacobs, 1980).

There are many causes for this poor technical performance. In some cases it is simply inadequate technical skills, measured by any yardstick, on the part of those designing, constructing and operating water points; in other cases the professionals, although well-trained and motivated, have too little local hydrological or geological information, or there simply has not yet been adequate local experience with different kinds of equipment or structures for sound designs to be drawn up. In such circumstances risks have to be taken and inevitably the failure rate is high. In some cases information might have been available if adequate administrative sources had existed to retrieve or extract it from archives or technical literature. In other cases quite inappropriate techniques and equipment have been used, either from a misplaced faith in the most recent technology or, far more often, because development was financed from external aid funds and these are all too often biased towards the use of complex technology imported from the country supplying the funds. Often complex imported equipment cannot cope with the local roads and no adequate system exists or can be quickly put in place for ordering, storing and supplying spare parts or other inputs.

5.4 Administrative and financial problems

Very often financial and administrative problems lie behind technical breakdowns. Broken equipment is not repaired because there is no money with which to do so, and even where there is, administrative procedures may act as glue in the bureaucratic machinery to prevent the work being done. It is not easy to collect watering fees in pastoral areas. For short periods at the height of the dry season very high rates of watering fees can be charged and collected, as the example from southeast Ethiopia quoted in Chapter 2 shows (Cossins, 1971a, p.44). But faced with high fees over an extended period, pastoralists will quickly move to alternative cheaper sources while these are available - only returning to the source charging a high fee at times of emergency. This makes the steady recovery of overhead costs for water development particularly difficult. Collecting fees for watering livestock is difficult for both private entrepreneurs and government. The data in Table 8 show how much, in 1979-1980, livestock owners in Botswana actually paid to use water at other people's or government boreholes.

Table 8. Costs of providing water and fees paid at public and private boreholes in Botswana (1979-1980).

Source: Bailey (1980, Tables 11 and 26).

^a This assumes that, on average, one 'livestock unit'- equivalent to an adult bovine - drinks 0.85 m³ of water per month throughout the year. This figure can be justified from Bailey (1980), Table 13.

^b 1 Pula = approx. US$ 0.90.

^c All figures are averages for many boreholes, and capital charges are at 'replacement' not 'historic' cost and include 12% interest on capital.

The data in Table 8 show how very much less than the average cost of providing water was paid by these livestock owners. In the case of privately-owned boreholes the average amount paid is almost exactly the same as the cost of the average amount of fuel used - an approximation to the marginal cost of water supply. Ties of friendship, fear of offending an important person, a desire to help the poor, ignorance of what the cost of supplying water actually is, all combine to make it difficult to charge the full-cost rate. In the case of both governments and private entrepreneurs not only are the rates for watering fees set too low, but even those set are not collected. Fortmann and Roe (1981, p. 24), for example, recount how they counted an average of about 300 cattle per day watering at a government borehole in a month where the total receipts for watering indicated only about 100 cattle watered; given a 2-day watering interval, this means that only about one fifth of the fees due were collected.

In many cases government financial procedures stipulate that money collected at a particular water point is not earmarked to meet the expenses incurred at that point but is paid into general government revenue and the expenses are met by a separate allocation. As a consequence users of the water point have absolutely no incentive to pay the fees, nor the employee operating the water point to collect them. The result of all these factors is that watering points in dry regions are constantly starved of funds for their operation, maintenance and repair. Such funds as are available in the responsible ministry for running water supplies will be allocated to regions of greater political priority.

Shortages of funds to run government water points are often compounded by administrative procedures for procurement and contracting. For example, in one African country funds were avail able to repair government open wells scattered over a huge area. Official procedures, however, required that the contract for the repair be awarded only after the receipt of three competitive tenders for the work. No contractor would tender for the work without inspecting each well to see precisely what repairs were required. But travailing round the wells and entering each would account for, by far, the largest part of the actual cost of repair, and for three contractors to be required to do this and for only one (the successful tenderer) to be paid, and then only for a subsequent trip on which to do the work, guaranteed that three bids were not received.

This is an extreme case but nicely illustrates the point that government procedures drawn up to fit the needs of a different set of circumstances are quite inappropriate for running livestock water supplies in dry areas. As a consequence operators are not paid their salaries, fuel for pumping runs out and is not replaced, and essential repairs and maintenance do not take place. The results are most dramatic in the case of boreholes because they either function or they do not. In the case of dams and hafirs the result is more usually a steady diminution in effectiveness over a period of time rather than the kind of instant catastrophe recorded above for the Tanzanian Maasai Range Project's dams.

5.5 Environmental problems and land use

The negative impact of water development on the environment has often been stressed in recent years. It is the counterpart to the increased livestock populations, and consequent increased livestock output already referred to. The 'last word' on the environmental impact of water points on African rangelands is still a long way from being written. An initial unrealistic optimism assumed that new water points would spread a finite grazing pressure more evenly and more widely; in practice the grazing pressure increased, in total, to fill the new niches made available by water development. The subsequent ultra-alarmist fears about the rapid spread of barren deserts around water points are now being replaced by more cautious assessments, as detailed attempts to measure change are revealing the complexity of the interactions between new water supplies, grazing pressure, climate, soil nutrient status, and the composition and productivity of the range vegetation.

The subject is beyond the scope of this study (for some reviews see Bernus, 1981, pp. 437-451; Breman et al, 1979/80; Horowitz, 1979, pp. 27-42; Warren and Maizels, 1977) except to note that the ill effects of new livestock water supplies on the range vegetation are more local in space, more circumscribed in time and less certain than has often been claimed. The timing, in the year, of the use of vegetation is often more important than the degree of grazing pressure. However, whereas previously the main concern was with the physical effects of trampling and grazing by animals on the soil structure and vegetation, more recently there has been a shift towards greater emphasis on the chemical effects of overgrazing on soil nutrient status (Penning de Vries and Djiteye, 1982). With this change in emphasis on the mechanism for degradation, it seems possible that new water supplies will be accorded a less prominent role.

Even if the environmental effects of new water supplies are still uncertain they have, nevertheless, had important effects on land-use systems. In part this is because, as intended, they reduced the distance livestock had to trek to water; and in part because they disrupted traditional patterns of management and control. Sometimes the new supplies are built by entrepreneurs, often with capital generated from outside the pastoral society, and these entrepreneurs' commercial and political contacts with non-pastoral society enable them to evade traditional social pressures, which might have made the use of the new water point conform to the traditional pattern. In most cases (Botswana presents an exception) the new supplies in dry regions 'belong' to government, i.e. government overtly claims the right to control access. Inevitably this means that additional people who would otherwise have been excluded from water points giving access to certain grazing areas are now allowed in.

In Botswana the number of livestock watering in the dry season at government or group owned or managed water points is significantly larger, even after allowing for differences in type of water supplies, than at private ones (Fortmann and Roe, 1981, pp. 88-120). Interestingly, the condition of the livestock at the private points is better while that of the grazing is worse, aptly illustrating the complexity of the issues involved. On government boreholes in Niger, Bernus has commented:

'To give good quality water to some and to refuse it to others has always been an insurmountable obstacle in principle and in practice. The choice is always difficult, because of the risk of favouring important chiefs and the richest pastoralists who reserve for themselves exclusive use of boreholes near their camp... The government... did not see how discriminatory regulation could be applied in such a vast area; neither did it want to run the risk of favouring some groups over others and causing disputes around the boreholes; for water is always the crucial point in rivalries between ethnic groups or between tribes.' (Bernus, 1977, p. 63).

In practice, although governments wish to evade the odium of such discrimination, failure to discriminate excites the resentment of the group that sees itself as being the rightful user of the area around the water point; and fighting between groups often forces a government's intervention in any case.

Changes in the pattern of control and management of water points have consequences for both efficiency and equity. In terms of efficiency, to the extent that control of access to water previously also regulated the grazing pressure, the governments' more relaxed attitudes to access are likely to increase this pressure. In many cases it also permits a change (and not always an environmentally or even, in the long term, an economically desirable change) in land use (see for example Fortmann and Roe, 1981, pp. 114-115). As a report on Tanzanian Maasailand puts it:

'And once a potentially stable, clean water supply is introduced into an area it almost automatically introduces a new population centre composed of representatives of modern society. At times (sic!) agriculturalists as well as outside pastoralists are attracted to the water.' (Hatfield and Ole Kuney, 1975/76, p. 18).

In many countries governments, at new government water points, are reluctant to enforce ethnic discrimination between members of on tribe or clan and another. The consequence is an increase in quarrels and fighting by those using the water point and a decline in the discipline of water use (Marty, 1972, pp. 33 and 43; Dahl and Sandford, 1978, p. 41). At the same time governments face a dilemma over the efficient management of water points. Efficient management, as we have seen in Chapter 2, requires rules about watering order (rosters), control of animals and cleaning up after they have drunk, and about day-to-day and less frequent maintenance. In some traditional societies these rules are adjudicated and enforced, with penalties where necessary, by 'well-masters', who bear ultimate responsibility to their own community. To decentralise such powers to a government employee in charge of a water point, and who is not responsible to the community, is an invitation to corruption. Not to decentralise but to retain the powers at some higher level in the official hierarchy may (or may not) reduce corruption but will certainly make management too remote to be effective.

5.6 The question of equity

Water development has equity as well as efficiency implications. By equity implications we mean the extent to which particular individuals or groups benefit relatively more than, or at the absolute expense of, others. We have already seen that some kinds of water source, i.e. open wells and many sandy river wells, require much more labour (per head) for watering livestock than do others, such as boreholes or dams. In circumstances where labour-extensive (i.e. non-intensive) types of water source abound, the demand for labour to water livestock will be low and the economic position of those whose only resource is their labour will be, in this respect, weak in relation to those with capital in the form of livestock. Cash wages will be low and non-cash institutions for redistributing livestock from labour-deficit to labour-surplus households relatively undeveloped or defunct. In contrast, in circumstances where labour-intensive types of water source abound, the position of those with much labour relative to capital in the form of livestock will be correspondingly strong. But demand for labour to water animals is only one source of demand for labour, and other sources of demand (e.g. for labour to work in the mines, as in Mauritania or Botswana) may be much more important. Given the general level of demand for labour, it may be advantageous for those with small herds below the size which can enjoy economies of scale in watering at labour-intensive sources (see earlier discussion in Chapter 3) to water these small herds at labour-extensive sources, thereby freeing as much as possible of their own labour for use in activities where returns are higher. We have already noted this phenomenon with boreholes in Niger (Marty, 1972, pp. 82 and 92) and with the avoidance of sandy river wells in Botswana (Fortmann and Roe, 1981, p. 73).

There is another aspect of water technology with equity implications. Traditional water sources, using resources and skills either endogenous to the society concerned or obtainable fairly easily from neighbouring societies, do not provide many opportunities for individuals to use their technical knowhow or their key positions in a technical post as a source of monopoly power or profit. In contrast, some kinds of modern water sources, particularly boreholes but also, to a lesser extent, large dams and-hafirs, do provide such opportunities. Although livestock owners may coerce borehole operators into opening up boreholes at times when they are not supposed to (Eddy, 1979, p. 168), conversely borehole operators in their turn are often able to extort bribes for doing the job for which they received a salary in any case. Their position as 'the men with the keys', either literally with respect to the engine or fuel store, or metaphorically as the only people who know how to operate the pumps, gives them considerable power. In Mali,

'The drilling of wells would not lessen the pressure on the ponds. Furthermore there is the problem of pulley devices often being controlled by security officers (guards). Quite frequently nomads cannot gain access to those 'monopolised' wells without resorting to threats or graft... Social control over access to subterranean waters is a real problem.' (Bourgeot, 1981, p. 174).

This problem of social control over modern technology also exists at a higher level. Where development programmes are heavily dependent on a particular form of technology for water supplies, the key technicians, the hydrologists, the drillers and the engineers in charge of heavy earth-moving equipment, are able to command salaries and influence out of all proportion to what they can obtain in an economy in which their skills are not in such short supply and the technology which they command is not such a key element. In Botswana, Kenya, Sudan and Tanzania, in theory it is the community, the politicians, the project managers and the land-use planners who control where and when water development takes place; in practice it is the technician in charge of the survey team or of the machine, whose judgement about what is feasible nobody else can challenge²³.

23 There is a discussion of this issue in Shepherd (1981).

Technology is important in a further and unexpected way. Water development is so popular in dry regions because it is so important to pastoralists' way of life. Those who control water points, be they private individuals or communities, reap monopoly profits from doing so, and when new supplies are developed the position of these monopolists is threatened. Some modern forms of water points are particularly susceptible to sabotage. In Niger boreholes were rendered useless by pastoralists who had thrown stones into the tubing (Bernus, 1977, p. 56); the same thing happened in southeast Ethiopia and to pipelines in Tanzanian Maasailand (source: discussions with range management officials).

The location, the density, the management and control of water points all have equity implications. Locating a new water point in an area that belongs to one group (A) of people rather than to another (B) may either benefit group A - because it gives them better access to water than group B or it may harm them because the livestock of group B now invade group A's grazing land since the new water point is open to all. It is very difficult to predict in advance what will happen. Livestock owners who demanded and welcomed a new water point in their area may subsequently come to regret this bitterly when they find that the assumptions they made that they would have privileged access turn out to be unfounded. If the density of water points is increased, this may help those whose households suffer a shortage of labour because, for example, they no longer have to decide whether to grow crops on the farm or herd livestock around a distant water point; with closer water points they can do both. Increasing the density of water points favours cattle, sheep and goats at the expense of camels, by opening up to grazing by all stock areas which previously only camels could reach; hence it also favours those people-ethnic groups or members of the family - who specialise in cattle rather than in camel husbandry.

Obviously the system of control over access to, and management of, a water point - including the system for deciding on its location and technology - has important implications for equity as well as efficiency. One can categorise systems in terms of the degree of centralization or decentralization of decision-making and of the extent to which users or operators of water points participate in this; or in terms of management style, whether it is authoritarian, liberal or contractual; and so on. Especially in dry areas, where access to water often also determines access to grazing, it is extremely important to potential users of a water point what system is adopted - for this in turn will influence, for example, the criteria by which eligibility to use a particular water point are decided, who adjudicates whether a particular person meets these criteria, and on what conditions, as to payment or participation in the labour of operation or maintenance, use is permitted²⁴. Procedures are also important; the procedures, for example, for claiming rights to water or for determining the order of watering on a particular day. Different systems will deal with these things in different ways. Consciously or unconsciously when new water points are developed choices are made which set up particular systems for management and control.

24 For further discussion of these issues, see Sandford (1983), especially Chapters 3, 4, 6 and 7.

There is not space here to explore all the implications for equity of different systems of management and control, nor to establish precise guidelines for different circumstances. The appropriate system will depend partly on whom one wants to be 'fair' to or to favour (e.g. the poorest, a particular tribe, women), partly on the local political and social structures, and partly on the local physical environment which may determine, for example, whether particular kinds of participatory institution are viable. In one part of Niger government ownership of new boreholes worked to the disadvantage of local Tuareg, because it allowed access to water (and so to grazing) to Fulani pastoralists in an area from which they had previously been largely excluded (Eddy, 1979, p.383). In Botswana, even before the adoption of the Tribal Grazing Lands Policy, which favours exclusive rights being granted to individuals to graze some areas, a policy of favouring ownership of new boreholes by private individuals or syndicates, and which even allowed the transfer of ownership of some existing boreholes from public to private hands (Hitchcock, 1979, p. 192), has

'Opened up new grazing and access to those who could afford the fees. Those who could not afford the fees had to remove their cattle to 'free' water supplies which were surrounded by heavily used grazing areas. The borehole owners in the process of drilling boreholes do become de facto owners of the grazing land surrounding them.' (Fortmann and Roe, 1981, p. 63 quoting Peters, 1980).

In northeast Kenya the fact that the 'grazing committees of pastoralists' who are supposed to advise government, inter alia, on the management of water supplies, conduct their meetings in Swahili - a language not understood by most pastoralists - and that attendance at these meetings is not paid, means that the 'pastoralists' who attend the meetings are predominantly traders in adjacent townships whose advice is heavily slanted towards their own interests (Helland, 1980, pp. 136-169; and the author's own observations).